MATLAB: Resulting neural network function

deep learningfitnetMATLABnftooltrain

How can the function and matrix of a neural network be retieved after training? After using the 'train' function, how can I export the matrix from the net object?

Best Answer

Use genFunction.

Related Solutions

MATLAB: How to use an M-by-N matrix input or an M-by-N matrix target in the NNTOOL GUI from the Neural Network Toolbox

The following example demonstrates how to do this.

At the MATLAB command prompt, type:

PNEW=rand(11,45); %You have 11 inputs to your network
TNEW=rand(5,45);% You have 5 targets to your network
nntool %starts the Neural Network gui

In the Neural Network GUI:

1. Import the data PNEW as 'Inputs' and TNEW as 'Targets' by using the Import button.

2. Create a new Network with the following information:

 Network Type:               Feed-Forward backprop
 Input ranges:               select PNEW from the dropdown menu
 Training Function:          TRAINLM
 Adaption learning function: LEARNGDM
 Performance function: MSE
 Number of Layers:           2

For the Properties of Layer 2, the number of neurons should be 5. The transfer function can be TANSIG.

3. Click on View for the network you created.

4. In the Training Info under the Train tab, select the Inputs as PNEW and Targets are TNEW

5. Click on 'Train Network'

MATLAB: Get matlab nural network parameter after training

The objects net and tr contain all of the parameters needed for training and evaluation The command

 net = net

leads to recovering all of the pretraining parameters.

The command

 tr = tr

leads to recovering all of the posttraining parameters and results.

Thank you for formally accepting my answer

Greg

Best Answer

Related Solutions

MATLAB: How to use an M-by-N matrix input or an M-by-N matrix target in the NNTOOL GUI from the Neural Network Toolbox

MATLAB: Get matlab nural network parameter after training

Related Question